two rural communities not far from Bangkok, 385 women aged 35 to 44 years reported a total of 1,121 previous live births; 447 (39.9 percent) were of birth rank five or higher. This does not settle the matter, of course; what is needed are studies in which both maternal age and parity are properly controlled. In the meantime, Perkin's contention (54) that older, high parity women constitute a "high risk" group seems reasonable. Knowledge, Attitude, and Practice (K.A.P.) surveys (15, 58) have shown that 85 percent of Thai women over 35 years of age, regardless of parity or number of living children, state that they want no more children. If these women could achieve that desire, one might expect to eliminate around a third of the maternal deaths. THE EVIDENCE CONCERNING HEALTH CONSEQUENCES OF BIRTH INTERVAL Although the total number of children occurring in a family is the most important determinant of population pressure within that family, the interval between births is a factor which must not be ignored. Obviously, in the absence of effective family planning, the mother who has children at frequent intervals is likely to have more of them. Equally obvious is the fact that the more young children there are in a household, the greater the demands upon the mother's energy and skills in providing adequate care for them. Interestingly enough, data relating infant mortality to birth interval have been available for decades. In 1923 Huse published a study carried out in Gary, Indiana, of 1,135 births, excluding first births (59). She found an infant mortality rate of 169.1 (per 1,000 live births) when the interval since the preceding child was less than 15 months, and the rate was 102.8 if the interval was greater than 24 months. Not long thereafter in 1925 Woodbury published the results of an investigation of some causal factors of infant mortality and included a study of the association between birth interval and infant mortality among 8,196 births:|! occurring in Baltimore, Maryland (60). He found the following variations: Birth interval 1 Year 2 Years 3 Years 4 Years Neonatal (up to 28 days) mortality rates 51.2 37.3 36.7 38.1 Infant mortality (up to 1 year) rates 146.7 98.6 86.5 84.9 Eastman, in discussing the effect of birth intervals 20 years later (62), quoted Woodbury's conclusions: *Yerushalmy (61) reports this figure as 7,882, but quotes the same mortality figures as Eastman. I have not personally been able to obtain Woodbury's original monograph.) have stated tha the effects of these two factors on increasing maternal risk are additive—eacl increases the risk independently.